Despite the advances in therapy of cancer, patients with localized (stage II or III) esophageal carcinoma when treated with preoperative chemoradiation have a dismal prognosis (<20% 5-year survival rate). Treatment is associated with dire consequences. Currently, none of the clinical or therapeutic parameters can help to individualize therapy (select certain treatments or avoid others). It is known that response to chemoradiation (as judged by examining the resected specimen) varies. Pathologic complete response (pathCR), defined as the absence of residual cancer, or lack of response (=50% residual cancer or exCRTR) dictates patient outcome. Approximately 25% of patients achieve a pathCR and another 25% exhibit exCRTR. Patients with pathCR survive much longer than those with <pathCR and patients with exCRTR live much shorter than those with <exCRTR. If one could identify biomarkers that would distinguish these different outcomes before patients received treatment, we could begin to individualize therapy (avoid surgery in pathCR patients and avoid chemoradiation in exCRTR patients). We hypothesize that certain biomarker signatures may highly correlate with pathCR or exCRTR. Thus, the specific aims for R21 are (1) Establish the distribution and clinical relevance (pathCR and exCRTR) of NF-kB, Gli-1, and SHH in esophageal cancer patients undergoing preoperative chemoradiation;(2A) Assess the effect of specific classes of drugs used on clinical outcome and (2 B) Establish quantitative requirements for the expression of selected drug-resistance biomarkers, establish independent clinical relevance and then study if they improve the specificity of the emerging NF-kB, Gli-1, and SHH signature(s) (3) Establish IHC standardized quantitative assays of SHH, Gli1, NF-kB and selected drug- related biomarkers in our CLIA certified laboratory. Upon reaching the prespecified milestones, we will embark on R33 with the following specific aims (1) Prospectively validate the established NF-kB, SHH, Gli-1, and drug- resistance biomarkers'signature(s) for pathCR and exCRTR from the retrospective cohort (60 preliminary + 175 retrospective = 235) in 212 samples from prospective patients who will have preoperative chemoradiation. (2A) Establish a predictive model for various clinical outcomes (pathCR, partial response [1% to 50% residual cancer], and exCRTR) in 447 patients (derived from the preliminary, retrospective, and prospective cohorts) to derive the most promising biomarker signature(s). (2B) Add clinical parameters and classes of drug to the nomogram established in specific aim 2A. Thus, this proposal seeks to pave the path to individualized therapy for esophageal cancer using biomarker signatures. PUBLIC HEALTH RELEVANCE: The goal of this study (R21/R33) is to study untreated cancer tissue from esophageal cancer patients receiving preoperative chemoradiation to discover if certain molecular biomarker classifiers can help individualize therapy. This is being proposed in two phases: retrospective (R21) and prospective (R33). Only promising findings in the R21 will be validated in the R33 phase. In the R21 phase, the immunohistochemistry (IHC) of NFkB, Shh and Gli-1 as well as relevant drug-related biomarkers will be used to establish a distribution and clinical relevance (pathCR and exCRTR) of biomarker signatures. The IHC methods will be standardized and transferred from the research laboratory to a certified IHC clinical laboratory.